Hello,
I'm just now getting into creating PCB's as a hobby, and created a little board that simply consists of a BLE module powered by a tiny Lithium polymer battery (40mah). Since the max operating voltage of the module is 3.6V, I also have an LDO that will drop the voltage down to 3.3V.

Everything works fine with the PCB's; The only problem I am having is that I will occasionally experience a crash of the LDO whenever the Lipo battery reaches max charge capacity during a charge cycle. As soon as the battery hits ~4.21 volts, and the Micro USB remains plugged in, the LDO will crash and its output voltage will spike from 3.3V to ~4.27 volts.

The only odd thing that happened to the board, is that the Schottky diode I will show you in the schematic below (D1), originally came assembled the wrong way around. And It wasn't until I powered the boards for a bit that I noticed this, and eventually de-soldered and put them back on the right way; But they did run a bit and charge with the diode the wrong way. Don't know if this would even matter though.

Whenever the LDO doesn't crash while the USB is still plugged in with a fully charged lipo, I can cause it to crash by testing the VBATT pin of the Charging-IC (U1) with my multi-meter. As soon as i touch my positive lead to that pin, the LDO's output voltage will crash. Regardless, the main point of this post is that it will sometimes crash without me touching the board at all.

Here is the power management portion of my schematic, Let me know if I've made any stupid mistakes that would cause this crash I described:




Note that LDO(U2) is: S-1200B33-A6T1U
And D1 is: BAT20JFILM

 

Is your mosfet (Q1) correct? Here is the diagram for a P-channel mosfet from Diodes Inc.:

 

Yes, here is an image of the board:



G is +5V from USB input.

Just can't seem to figure out why the LDO is crashing

 

Have you tried putting a oscilloscope on the output of the MCP73831 when the battery is charged? specifically if the 73831 was oscillating on and off.

Taking Q1 out. and connecting per 73831 data sheet Load connected to battery. With a current sufficient to run circuit and charge lipo has worked for me.

 

Read the 'precautions' section of the datasheet.
What is the load current of the LDO when it spikes?
Is it just a spike or does it stay at 4.2V?

 

It looks like we used incorrect bypass capacitor values for the lipo charging IC.
The datasheet recommended 4uF caps, so 12pF was obviously a mistake. I believe this will solve the issue when the source switches to constant voltage mode after the battery finishes charging. At least I'm hoping this is the solution